[HN Gopher] How University Students Use Claude
___________________________________________________________________
How University Students Use Claude
Author : pseudolus
Score : 164 points
Date : 2025-04-09 15:41 UTC (7 hours ago)
(HTM) web link (www.anthropic.com)
(TXT) w3m dump (www.anthropic.com)
| stv_123 wrote:
| Interesting article, but I think it downplays the incidence of
| students using Claude as an alternative to building foundational
| skills. I could easily see conversations that they outline as
| "Collaborative" primarily being a user walking Claude through
| multi-part problems or asking it to produce justifications for
| answers that students add to assignments.
| yieldcrv wrote:
| > I think it downplays the incidence of students using Claude
| as an alternative to building foundational skills
|
| I think people will get more utility out of education programs
| that allow them to be productive with AI, at the expense of
| foundational knowledge
|
| Universities have a different purpose and are tone deaf to why
| their students use universities for the last century: which is
| that the corporate sector decided university degrees were
| necessary despite 90% of the cross disciplinary learning being
| irrelevant.
|
| Its not the university's problem and they will outlive this
| meme of catering to the middle class' upwards mobility at all.
| They existed before and will exist after.
|
| The university may never be the place for a human to hone the
| skill of being augmented with AI but a trade school or bootcamp
| or other structured learning environment will be, for those not
| self started enough to sit through youtube videos and trawl
| discord servers
| fallinditch wrote:
| Yes, AI tools have shifted the education paradigm and
| cognition requirements. This is a 'threat' to universities,
| but I would also argue that it's an opportunity for
| universities to reinvent the experience of further education.
| ryandrake wrote:
| Yea, the solution here is to embrace the reality that these
| tools exist and will be used regardless of what the
| university wants, and use it as an opportunity to level
| _up_ the education and experience.
|
| The clueless educational institutions will simply try to
| fight it, like they tried to fight copy/pasting from Google
| and like they probably fought calculators.
| mppm wrote:
| > Interesting article, but I think it downplays the incidence
| of students using Claude as an alternative to building
| foundational skills.
|
| No shit. This is anecdotal evidence, but I was recently
| teaching a university CS class as a guest lecturer (at a
| somewhat below-average university), and almost all the students
| were basically copy-pasting task descriptions and error
| messages into ChatGPT in lieu of actually programming. No one
| seemed to even read the output, let alone be able to explain
| it. "Foundational skills" were near zero, as a result.
|
| Anyway, I strongly suspect that this report is based on careful
| whitewashing and would reveal 75% cheating if examined more
| closely. But maybe there is a bit of sampling bias at play as
| well -- maybe the laziest students just never bother with
| anything but ChatGPT and Google Colab, while students using
| Claude have a little more motivation to learn something.
| colonial wrote:
| CS/CE undergrad here who entered university right when
| ChatGPT hit. Things are _bad_ at my large state school.
|
| People who spent the past two years offloading their entry-
| level work onto LLMs are now taking 400-level systems
| programming courses and running face-first into a capability
| wall. I try my best to help, but there's only so much I can
| do when basic concepts like structs and pointer manipulation
| get blank stares.
|
| > "Oh, the foo field in that struct should be signed instead
| of unsigned."
|
| < "Struct?"
|
| > "Yeah, the type definition of Bar? It's right there."
|
| < "Man, I had ChatGPT write this code."
|
| > "..."
| jjmarr wrote:
| Put the systems level programming in year 1, honestly.
| Either you know the material going in, or you fail out.
| tmpz22 wrote:
| Direct quote I heard from an undergrad taking statistics:
|
| "Snapchat AI couldn't get it right so I skipped the assignment"
| dvngnt_ wrote:
| back in my day we used snap to send spicy photos now they're
| using AI to cheat on homework. im not sure what's worse
| moffkalast wrote:
| Well if statistics can't understand itself, then what hope do
| the rest of us have?
| ilrwbwrkhv wrote:
| AI bubble seems close to collapsing. God knows how many billions
| have been invested and we still don't have an actual use case for
| AI which is good for humanity.
| boredemployee wrote:
| I think I understand what you're trying to say.
|
| We certainly improve productivity, but that is not necessarily
| good for humanity. Could be even worse.
|
| i.e.: my company already expect less time for some tasks given
| that they _know_ I'll probably use some AI to do tasks. Which
| means I can humanly handle more context in a given week if the
| metric is "labour", but you end up with your brain completely
| melted.
| bluefirebrand wrote:
| > We certainly improve productivity
|
| I think this is really still up for debate
|
| We produce more output certainly but if it's overall lower
| quality than previous output is that really "improved
| productivity"?
|
| There has to be a tipping point somewhere, where faster
| output of low quality work is actually decreasing
| productivity due to the efforts now required to keep the
| tower of garbage from toppling
| fourseventy wrote:
| It's not up for debate. Ask any programmer if LLMs improve
| productivity and the answer is 100% yes.
| AlexandrB wrote:
| Meanwhile in this article/thread you have a bunch of
| programmers complaining that LLMs don't improve overall
| productivity:
| https://news.ycombinator.com/item?id=43633288
| bluefirebrand wrote:
| I am a programmer and my opinion is that all of the AI
| tooling my company is making me use gets in the way about
| as often as it helps. It's probably overall a net
| negative, because any code it produces for me takes
| longer for me to review and ensure correctness as it
| would to just write it
|
| Does my opinion count?
| DickingAround wrote:
| I think the core of the 'improved productivity' question will
| be ultimately impossible to answer. We would want to know if
| productivity was improved over the lifetime of a society;
| perhaps hundreds of years. We will have no clear A/B test
| from which to draw causal relationships.
| AlexandrB wrote:
| This is exactly right. It also depends on how all the AGI
| promises shake out. If AGI really does emerge soon, it
| might not matter anymore whether students have any
| foundational knowledge. On the other hand, if you still
| need people to know stuff in the future, we might be
| creating a generation of citizens incapable of doing the
| job. That could be catastrophic in the long term.
| amiantos wrote:
| Your statement appears to be composed almost entirely of vague
| and ambiguous statements.
|
| "AI bubble seems close to collapsing" in response to an article
| about AI being used as a study aid. Does not seem relevant to
| the actual content of the post at all, and you do not provide
| any proof or explanation for this statement.
|
| "God knows how many billions have been invested", I am pretty
| sure it's actually not that difficult to figure out how much
| investor money has been poured into AI, and this still seems
| totally irrelevant to a blog post about AI being used as a
| study aid. Humans 'pour' billions of dollars into all sorts of
| things, some of which don't work out. What's the suggestion
| here, that all the money was wasted? Do you have evidence of
| that?
|
| "We still don't have an actual use case for AI which is good
| for humanity"... What? We have a lot of use cases for AI, some
| of which are good for humanity. Like, perhaps, as a study aid.
|
| Are you just typing random sentences into the HN comment box
| every time you are triggered by the mention of AI? Your post is
| nonsense.
| papichulo2023 wrote:
| It is helping me do that projects that would otherwise take me
| hours in just a few minutes, soooo, shrug.
| user432678 wrote:
| What kind of projects are those? I am genuinely curious. I
| was excited by AI, Claude specifically, since I am an avid
| procrastinator and would love to finish tens of projects I
| have in mind. Most of those projects are games with
| specifical constraints. I got disenchanted pretty quickly
| when started actually using AI to help with different parts
| of the game programming. Majority of problems I had are
| related to poor understanding of generated code. I mean yes,
| I read the code, fixed minor issues, but it always feels like
| I don't really internalised the parts of the game which slows
| me down quite significantly in a long run, when I need to
| plan major changes. Probably a skill issue, but for now the
| only thing AI is helpful for me is populating Jira
| descriptions for my "big picture refactoring" work. That's
| basically it.
| noman-land wrote:
| I was able to use llama.cpp and whisper.cpp to help me
| build a transcription site for my favorite podcast[0]. I'm
| a total python noob and hadn't really used sqlite before,
| or really used AI before but using these tools, _completely
| offline_ , llama.cpp helped me write a bunch of python and
| sql to get the job done. It was incredibly fun and
| rewarding and most importantly, it got rid of the dread of
| not knowing.
|
| 0 - https://transcript.fish
| SamBam wrote:
| I feel like Anthropic has an incentive to minimize how much
| students use LLMs to write their papers for them.
|
| In the article, I guess this would be buried in
|
| > Students also frequently used Claude to provide technical
| explanations or solutions for academic assignments
| (33.5%)--working with AI to debug and fix errors in coding
| assignments, implement programming algorithms and data
| structures, and explain or solve mathematical problems.
|
| "Write my essay" would be considered a "solution for academic
| assignment," but by only referring to it obliquely in that
| paragraph they don't really tell us the prevalence of it.
|
| (I also wonder if students are smart, and may keep outright usage
| of LLMs to complete assignments on a separate, non-university
| account, not trusting that Anthropic will keep their
| conversations private from the university if asked.)
| radioactivist wrote:
| Most of their categories have straightforward interpretations
| in terms of students using the tool to cheat. They don't seem
| to want to/care to analyze that further and determine which are
| really cheating and which are more productive uses.
|
| I think that's a bit telling on their motivations (esp. given
| their recent large institutional deals with universities).
| SamBam wrote:
| Indeed. I called out the second-top category, but you could
| look at the top category as well:
|
| > We found that students primarily use Claude to create and
| improve educational content across disciplines (39.3% of
| conversations). This often entailed designing practice
| questions, editing essays, or summarizing academic material.
|
| Sure, throwing a paragraph of an essay at Claude and asking
| it to turn it into a 3-page essay could have been categorized
| as "editing" the essay.
|
| And it seems pretty naked the way they lump "editing an
| essay" in with "designing practice questions," which are
| clearly very different uses, even in the most generous
| interpretation.
|
| I'm not saying that the vast majority of students _do_ use AI
| to cheat, but I do want to say that, if they _did_ , you
| could probably write this exact same article and tell no
| lies, and simply sweep all the cheating under titles like
| "create and improve educational content."
| vunderba wrote:
| Exactly. There's a big difference between a student having a
| back-and-forth dialogue with Claude around _" the extent to
| which feudalism was one of the causes of the French
| Revolution."_, versus another student using their smartphone to
| take a snapshot of the actual homework assignment, pasting it
| into Claude and calling it a day.
| PeterStuer wrote:
| From what I could observe, the latter is endemic amongst high
| school students. And don't kid yourself. For many it is just
| a step up from copy/pasting the first Google result.
|
| They never could be arsed to learn how to input their
| assignments into Wolfram Alpha. It was always the ux/ui
| effort that held them back.
| j2kun wrote:
| They use an LLM to summarize the chats, which IMO makes the
| results as fundamentally unreliable as LLMs are. Maybe for an
| aggregate statistical analysis (for the purpose of...vibe-based
| product direction?) this is good enough, but if you were to use
| this to try to inform impactful policies, caveat emptor.
| j2kun wrote:
| For example, it's fashionable in math education these days to
| ask students to generate problems as a different mode of
| probing understanding of a topic. And from the article: "We
| found that students primarily use Claude to create and improve
| educational content across disciplines (39.3% of
| conversations). This often entailed designing practice
| questions, ..." That last part smells fishy, and even if you
| saw a prompt like "design a practice question..." you wouldn't
| be able to know if they were cheating, given the context
| mentioned above.
| dtnewman wrote:
| > A common question is: "how much are students using AI to
| cheat?" That's hard to answer, especially as we don't know the
| specific educational context where each of Claude's responses is
| being used.
|
| I built a popular product that helps teachers with this problem.
|
| Yes, it's "hard to answer", but let's be honest... it's a very
| very widespread problem. I've talked to hundreds of teachers
| about this and it's a ubiquitous issue. For many students, it's
| literally "let me paste the assignment into ChatGPT and see what
| it spits out, change a few words and submit that".
|
| I think the issue is that it's so tempting to lean on AI. I
| remember long nights struggling to implement complex data
| structures in CS classes. I'd work on something for an hour
| before I'd have an epiphany and figure out what was wrong. But
| that struggling was ultimately necessary to really learn the
| concepts. With AI, I can simply copy/paste my code and say "hey,
| what's wrong with this code?" and it'll often spot it (nevermind
| the fact that I can just ask ChatGPT "create a b-tree in C" and
| it'll do it). That's amazing in a sense, but also hurts the
| learning process.
| stv_123 wrote:
| Yeah, the concept of "productive struggle" is important to the
| education process and having a way to short circuit it seems
| like it leads to worse learning outcomes.
| umpalumpaaa wrote:
| I am not sure all humans work the same way though. Some get
| very very nervous when they begin to struggle. So nervous
| that they just stop functioning.
|
| I felt that during my time in university. I absolutely loved
| reading and working through dense math text books but the
| moment there was a time constraint the struggle turned into
| chaos.
| AlexandrB wrote:
| > Some get very very nervous when they begin to struggle.
| So nervous that they just stop functioning.
|
| I sympathize, but it's impossible to remove all struggle
| from life. It's better in the long run to work through this
| than try to avoid it.
| yapyap wrote:
| > I think the issue is that it's so tempting to lean on AI. I
| remember long nights struggling to implement complex data
| structures in CS classes. I'd work on something for an hour
| before I'd have an epiphany and figure out what was wrong. But
| that struggling was ultimately necessary to really learn the
| concepts. With AI, I can simply copy/paste my code and say
| "hey, what's wrong with this code?" and it'll often spot it
| (nevermind the fact that I can just ask ChatGPT "create a
| b-tree in C" and it'll do it). That's amazing in a sense, but
| also hurts the learning process.
|
| In the end the willingness to struggle will set apart the truly
| great Software Engineer from the AI-crutched. Now of course
| this will most of the time not be rewarded, when a company
| looks at two people and sees "passable" code from both but one
| is way more "productive" with it (the AI-crutched engineer)
| they'll inititally appreciate this one more.
|
| But in the long run they won't be able to explain the choices
| made when creating the software, we will see the retraction
| from this type of coding when the first few companies' security
| falls apart like a house of cards due to AI reliance.
|
| It's basically the "instant gratification vs delayed
| gratification" argument but wrapped in the software dev box.
| JohnMakin wrote:
| I don't wholly disagree with this post, but I'd like to add a
| caveat, observing my own workflow with these tools.
|
| I guess I'd qualify to you as someone "AI crutched" but I
| mostly use it for research and bouncing ideas (or code
| complete, which I've mentioned before - this is a great use
| of the tool and I wouldn't consider it a crutch, personally).
|
| For instance, "parse this massive log output, and highlight
| anything interesting you see or any areas that may be a
| problem, and give me your theories."
|
| Lots of times its wrong. Sometimes its right. Sometimes, its
| response gives me an idea that leads to another direction.
| It's essentially how I was using google + stack overflow ten
| years ago - see your list of answers, use your intuition,
| knowledge, and expertise to find the one most applicable to
| you, continue.
|
| This "crutch" is essentially the same one I've always used,
| just in different form. I find it pretty good at doing code
| review for myself before I submit something more formal, to
| catch any embarrassing or glaringly obvious bugs or incorrect
| test cases. I would be wary of the dev that refused to use
| tools out of some principled stand like this, just as I'd be
| wary of a dev that overly relied on them. There is a balance.
|
| Now, if all you know are these tools and the workflow you
| described, yea, that's probably detrimental to growth.
| vunderba wrote:
| I've been calling this out since the rise of ChatGPT:
|
| "The real danger lies in their seductive nature - over how
| tempting it becomes to immediately reach for the LLM to provide
| an answer, rather than taking a few moments to quietly ponder
| the problem on your own. By reaching for it to solve any
| problem at nearly an instinctual level you are completely
| failing to cultivate an intrinsically valuable skill - that of
| critical reasoning."
| nonethewiser wrote:
| Somewhat agree.
|
| I agree in principal - the process of problem solving is the
| important part.
|
| However I think LLMs make you do more of this because of what
| you can offload to the LLM. You can offload the simpler
| things. But for the complex questions that cut across
| multiple domains and have a lot of ambiguity? You're still
| going to have to sit down and think about it. Maybe once
| you've broken it into sufficiently smaller problems you can
| use the LLM.
|
| If we're worried about abstract problem solving skills that
| doesnt really go away with better tools. It goes away when we
| arent the ones using the tools.
| Peritract wrote:
| You can offload the simpler things, but struggling with the
| simpler things is how you build the skills to handle the
| more complex ones that you can't hand off.
|
| If the simpler thing in question is a task you've already
| mastered, then you're not losing much by asking an LLM to
| help you with it. If it's not trivial to you though, then
| you're missing an opportunity to learn.
| jplusequalt wrote:
| Couldn't have said it better myself.
|
| The biology of the human brain will not change as a
| result of these LLMs. We are imperfect and will tend to
| take the easiest route in most cases. Having an "all
| powerful" tool that can offload the important work of
| figuring out tough problems seems like it will lead to a
| society less capable in solving complex problems.
| nonethewiser wrote:
| If you haven't mastered it yet then its not a simple
| thing.
|
| Grandma will not be able to implement a simple add
| function using python by asking chat gpt and copy
| pasting.
| bko wrote:
| When modern search became more available, a lot of people said
| there's no point of rote memorization as you can just do a
| Google search. That's more or less accepted today.
|
| Whenever we have a new technology there's a response "why do I
| need to learn X if I can always do Y", and more or less, it has
| proven true, although not immediately.
|
| For instance, I'm not too concerned about my child's ability to
| write very legibly (most writing is done on computers), spell
| very well (spell check keeps us professional), reading a map to
| get around (GPS), etc
|
| Not that these aren't noble things or worth doing, but they
| won't impact your life too much if you're not interest in
| penmanship, spelling, or cartography.
|
| I believe LLMs are different (I am still stuck in the moral
| panic phase), but I think my children will have a different
| perspective (similar to how I feel about memorizing poetry and
| languages without garbage collection). So how do I answer my
| child when he asks "Why should I learn to do X if I can just
| ask an LLM and it will do it better than me"
| nyeah wrote:
| Spell check isn't really adequate. You get a page full of
| correctly spelled words, but they're the wrong words.
| walthamstow wrote:
| Try being British, often they're not correctly spelt words
| at all.
| delusional wrote:
| "More or less" is doing a lot of work there. School, at least
| where I am, still spends the first year getting children to
| memorize the order of the numbers from 1-20 and if there's an
| even or odd number of a thing on a picture.
|
| Do you google if 5 is less than 6 or do you just memorize
| that?
|
| If you believe that creativity is not based on a foundation
| of memorization and experience (which is just memorization)
| you need to reflect on the connection between those.
| noitpmeder wrote:
| This is an insane take.
|
| The issue is that, when presented with a situation that
| requires writing legibly, spelling well, or reading a map,
| WITHOUT their AI assistants, they will fall apart.
|
| The AI becomes their brain, such that they cannot function
| without it.
|
| I'd never want to work with someone who is this reliant on
| technology.
| Vvector wrote:
| Do you have the skills and knowledge to survive like a
| pioneer from 200 years ago?
|
| Technology is rapidly changing humanity. Maybe for the
| worse.
| tux1968 wrote:
| Indeed. More people need to grow their own vegetables. AI
| may undermine our ability for high level abstract
| thought, but industrial agriculture already represents an
| existential threat, should it be interrupted for any
| reason.
| kibwen wrote:
| Knowledge itself is the least concern here. Human society
| is extremely good at transmitting information. More
| difficult to transmit are things like critical thinking
| and problem-solving ability. Developing meta-cognitive
| processes like the latter are the real utility of
| education.
| bko wrote:
| Maybe 40 years ago there were programmers that would not
| work with anyone that use IDEs or automated memory
| management. When presented with a programming task that
| requires these things and you're WITHOUT your IDE or
| whatever, they will fall apart.
|
| Look, I agree with you, I'm just trying to articulate to
| someone why they should learn X if they believe an LLM
| could help them and "an LLM won't always be around" isn't a
| good argument, because lets be honest, it likely will. This
| is the same thing as "you won't walk around all day with a
| calculator in your pocket so you need to learn math"
| Hasu wrote:
| > This is the same thing as "you won't walk around all
| day with a calculator in your pocket so you need to learn
| math"
|
| People who can't do simple addition and multiplication
| without a calculator (12*30 or 23 + 49) are absolutely at
| a disadvantage in many circumstances in real life and I
| don't see how you could think this isn't true. You can't
| work as a cashier without this skill. You can't play
| board games. You can't calculate tips or figure out how
| much you're about to spend at the grocery store. You
| could pull out your phone and use a calculator in all
| these situations, but people don't.
| dwaltrip wrote:
| You are also likely to be more vulnerable to financial
| mishaps and scams.
| gwervc wrote:
| A lot of developers of my generation (30+) learned to
| program within a code editor and compile their project in
| command line. Remove the IDE and we can still code.
|
| On the other hand my master 2 students, most of which
| learned scripting in the previous year, can't even split
| a project in multiple files after being explained
| multiple times. Some have more knowledge and ability than
| others, but a signifiant fraction is just copy-pasting
| LLM output to solve whatever is asked from them instead
| of trying to do it themselves, or asking questions.
| rurp wrote:
| I think the risk isn't just that LLMs won't exist, but
| that they will fail at certain tasks that need to get
| done. Someone who is highly dependent on prompt
| engineering and doesn't understand any of the underlying
| concepts is going to have a bad time with problems they
| can't prompt their way out of.
|
| This is something I see with other tools. Some people get
| highly dependent on things like advanced IDE features and
| don't care to learn how they actually work. That works
| fine most of the time but if they hit a subtle edge case
| they are dead in the water until someone else bails them
| out. In a complicated domain there are always edge cases
| out there waiting to throw a wrench in things.
| mbesto wrote:
| Do you work with people who can multiply 12.3% * 144,005.23
| rapidly without a calculator?
|
| > The issue is that, when presented with a situation that
| requires writing legibly, spelling well, or reading a map,
| WITHOUT their AI assistants, they will fall apart.
|
| The parent poster is positing that for 90% of cases they
| WILL have their AI assistant because its in their pocket,
| just like a calculator. It's not insane to think that and
| its a fair point to ponder.
| HDThoreaun wrote:
| It's all about critical thinking. The answer to your kid is
| that LLMs are a tool and until they run the entire economy
| there will still need to be people with critical thinking
| skills making decisions. Not every task at school helps hone
| critical thinking but many of them do.
| dingnuts wrote:
| > That's more or less accepted today.
|
| Bullshit! You cannot do second order reasoning with a set of
| facts or concepts that you have to look up first.
|
| Google Search made intuition and deep understanding and
| encyclopedic knowledge MORE important, not less.
|
| People will think you are a wizard if you read documentation
| and bother to remember it, because they're still busy asking
| Google or ChatGPT while you're happily coding without pausing
| vonneumannstan wrote:
| I am 100% certain people said the same thing about
| arithmetic and calculators and now mental arithmetic skill
| is nothing more than a curiosity.
| dwaltrip wrote:
| I encourage you to reconsider.
|
| Mental math is essential for having strong numerical
| fluency, for estimation, and for reasoning about many
| systems. Those skills are incredibly useful for thinking
| critically about the world.
| joe5150 wrote:
| Being able to do basic math in your head is valuable just
| in terms of basic practicality (quickly calculating a tip
| or splitting a bill, doubling a recipe, reasoning about a
| budget...), but this is a poor analogy anyway because 3x2
| is still 3x2 regardless of how you get there whereas
| creative work produced by software is worthless.
| joe5150 wrote:
| > Google Search made intuition and deep understanding and
| encyclopedic knowledge MORE important, not less.
|
| Not to mention discernment and info literacy when you do
| need to go to the web to search for things. AI content slop
| has put everybody who built these skills on the back foot
| again, of course.
| kibwen wrote:
| The irreducible answer to "why should I" is that it makes you
| ever-more-increasingly reliant on a teetering tower of
| fragile and interdependent supply chains furnished by for-
| profit companies who are all too eager to rake you over the
| coals to fulfill basic cognitive functions.
|
| Like, Socrates may have been against writing because he
| thought it made your memory weak, but at least I, an
| individual, am perfectly capable of manufacturing my own
| writing implements with a modest amount of manual labor and
| abundantly-available resources (carving into wood, burning
| wood into charcoal to write on stone, etc.). But I ain't
| perfectly capable of doing the same to manufacture an
| integrated circuit, let alone a digital calculator, let alone
| a GPU, let alone an LLM. Anyone who delegates their thought
| to a corporation is permanently hitching their fundamental
| ability to think to this wagon.
| notyourwork wrote:
| Although I agree, convincing children to learn using that
| rationalization won't work.
| bigstrat2003 wrote:
| Yes it does. Plenty of children accept "you won't always
| have (tool)" as a reason for learning.
| jplusequalt wrote:
| All adults were once children and there are plenty of
| adults who cannot read beyond a middle school reading
| level or balance a simple equation. This has been a
| problem before we ever gave them GPTs. It stands to
| reason it will only worsen in a future dominated by them.
| freeone3000 wrote:
| "You won't always have a calculator" became moderately
| false to laughably false as I went from middle to high
| school. Every task I will ever do for money will be done
| on a computer.
|
| I'm still garbage at arithmetic, especially mental math,
| and it really hasn't inhibited my career in any way.
| hackyhacky wrote:
| > The irreducible answer to "why should I" is that it makes
| you ever-more-increasingly reliant on a teetering tower of
| fragile and interdependent supply chains furnished by for-
| profit companies who are all too eager to rake you over the
| coals to fulfill basic cognitive functions.
|
| Yes, but that horse has long ago left the barn.
|
| I don't know how to grow crops, build a house, tend
| livestock, make clothes, weld metal, build a car, build a
| toaster, design a transistor, make an ASIC, or write an OS.
| I _do_ know how to write a web site. But if I cede that
| skill to an automated process, then _that_ is the feather
| that will break the camel 's back?
|
| The history of civilization is the history of
| specialization. No one can re-build all the tools they rely
| on from scratch. We either let other people specialize, or
| we let machines specialize. LLMs are one more step in the
| latter.
|
| The Luddites were right: the machinery in cotton mills was
| a direct threat to their livelihood, just as LLMs are now
| to us. But society marches on, textile work has been
| largely outsourced to machines, and the descendants of the
| Luddites are doctors and lawyers (and coders). 50 years
| from new the career of a "coder" will evoke the same
| historical quaintness as does "switchboard operator" or
| "wainwright."
| ryandrake wrote:
| This reply brings to mind the well-known Heinlein quote:
| A human being should be able to change a diaper, plan an
| invasion, butcher a hog, conn a ship, design a building,
| write a sonnet, balance accounts, build a wall, set a
| bone, comfort the dying, take orders, give orders,
| cooperate, act alone, solve equations, analyze a new
| problem, pitch manure, program a computer, cook a tasty
| meal, fight efficiently, die gallantly. Specialization is
| for insects.
| marcosdumay wrote:
| The sheer amount of activities that he left out because
| he couldn't even remember they existed would turn this
| paragraph into a book.
| DrillShopper wrote:
| This is a fantastic and underrated quote, despite all of
| the problems I have with Heinlein's fascism-glorifying
| work.
| crooked-v wrote:
| That's a quote that sounds great until, say, that self-
| built building by somebody who's neither engineer nor
| architect at best turns out to have some intractible
| design flaw and at worst collapses and kills people.
|
| It's also a quote from a character who's literally
| immortal and so has all the time in the world to learn
| things, which really undermines the premise.
| Mawr wrote:
| What an awful quote. Literally all progress we've made is
| due to ever increasing specialization.
| djhn wrote:
| It's a quote from a character in Heinlein's fiction. A
| human character with a lifespan of over a thousand years.
|
| I too liked that quote and found it inspiring. Until I
| read the book, that is.
| theLiminator wrote:
| I think removing pointless cognitive load makes sense,
| but the point of an education is to learn how to
| think/reason. Maybe if we get AGI there's no point
| learning that either, but it is definitely not great if
| we get a whole generation who skip learning how to
| problem solve/think due to using LLMs.
|
| IMO it's quite different than using a calculator or any
| other tool. It can currently completely replace the human
| in the loop, whereas with other tools they are generally
| just a step in the process.
| hackyhacky wrote:
| > IMO it's quite different than using a calculator or any
| other tool. It can currently completely replace the human
| in the loop, whereas with other tools they are generally
| just a step in the process.
|
| The (as yet unproven) argument for the use of AIs is that
| using AI to solve simpler problems allows us humans to
| focus on the big picture, in the same way that letting a
| calculator solve arithmetic gives us flexibility to
| understand the math behind the arithmetic.
|
| No one knows if that's true. We're running a grand
| experiment: the next generation will either surpass us in
| grand fashion using tools that we couldn't imagine, or
| will collapse into a puddle of ignorant consumerism, a la
| Wall-E.
| jplusequalt wrote:
| >the next generation will either surpass us in grand
| fashion using tools that we couldn't imagine, or will
| collapse into a puddle of ignorant consumerism, a la
| Wall-E
|
| Seeing how the world is based around consumerism, this
| future seems more likely.
|
| HOWEVER, we can still course correct. We need to
| organize, and get the hell off social media and the
| internet.
| hackyhacky wrote:
| > HOWEVER, we can still course correct. We need to
| organize, and get the hell off social media and the
| internet.
|
| Given what I know of human nature, this seems improbable.
| jplusequalt wrote:
| I think it's possible. I think the greatest trick our
| current societal structure ever managed to pull, is the
| proliferation of the belief that any alternatives are
| impossible. "Capitalist realism"
|
| People who organize tend to be the people who are most
| optimistic about change. This is for a reason.
| harikb wrote:
| It may be possible for you (I am assuming you are > 20,
| mature adult). But the context is around teens in the
| prime of their learning. It is too hard to keep
| ChatGPT/Claude away from them. Social media is too
| addictive. Those TikTok/Reels/Shorts are addictive and
| never ending. We are doomed imho.
|
| If education (schools) were to adopt a teaching-AI (one
| that will given them the solution, but at least ask a
| bunch of questions ), may be there is some hope.
| jplusequalt wrote:
| >We are doomed imho.
|
| I encourage you to take action to prove to yourself that
| real change is possible.
|
| What you can do in your own life to enact change is hard
| to say, given I know nothing about your situation. But
| say you are a parent, you have control over how often
| your children use their phones, whether they are on
| social media, whether they are using ChatGPT to get
| around doing their homework. How we raise the next
| generation of children will play an important role in how
| prepared they are to deal with the consequences of the
| actions we're currently making.
|
| As a worker you can try to organize to form a union. At
| the very least you can join an organization like the
| Democratic Socialists of America. Your ability to
| organize is your greatest strength.
| palmotea wrote:
| > The (as yet unproven) argument for the use of AIs is
| that using AI to solve simpler problems allows us humans
| to focus on the big picture, in the same way that letting
| a calculator solve arithmetic gives us flexibility to
| understand the math behind the arithmetic.
|
| And I can tell you _from experience_ that "letting a
| calculator solve arithmetic" (or more accurately, being
| dependent on a calculator to solve arithmetic) means you
| cripple your ability to learn and understand more
| advanced stuff. _At best_ your decision turned you into
| the equivalent of a computer trying to run a 1GB binary
| with 8MB of RAM and _a lot_ of paging.
|
| > No one knows if that's true. We're running a grand
| experiment: the next generation will either surpass us in
| grand fashion using tools that we couldn't imagine, or
| will collapse into a puddle of ignorant consumerism, a la
| Wall-E.
|
| It's the latter. Though I suspect the masses will be
| shoved into the garbage disposal than be allowed to
| wallow in ignorant consumerism. Only the elite that owns
| the means of production will be allowed to indulge.
| whilenot-dev wrote:
| I think the latest GenAI/LLM bubble shows that tech (this
| _hype_ kind of tech) doesn 't want us to learn, to think
| or reason. It doesn't want to be seen as a mere tool
| anymore, it wants to drive under the appearance that it
| can reason on its own. We're in the process where tech
| just wants us to adapt to it.
| tristor wrote:
| > I don't know how to grow crops, build a house, tend
| livestock, make clothes, weld metal, build a car, build a
| toaster, design a transistor, make an ASIC, or write an
| OS.
|
| Why not? I mean that, quite literally.
|
| I don't know how to make an ASIC, and if I tried to write
| an OS I'd probably fail miserably many times along the
| way but might be able to muddle through to something very
| basic. The rest of that list is certainly within my
| wheelhouse even though I've never done any of those
| things professionally.
|
| The peer commenter shared the Heinlein quote, but there's
| really something to be said for /society/ of being
| peopled by well-rounded individuals that are able to
| competently turn themselves to many types of tasks.
| Specialization can also be valuable, but specialization
| in your career should not prevent you from gaining a
| breadth of skills outside of the workplace.
|
| I don't know how to do any of the things in your list
| (including building a web site) as an /expert/, but it
| should not be out of the realm of possibility or even
| expectation that people should learn these things at the
| level of a competent amateur. I have grown a garden, I
| have worked on a farm for a brief time, I've helped build
| houses (Habitat for Humanity), I've taken a hobbyist
| welding class and made some garish metal sculptures, I've
| built a race car and raced it, and I've never built a
| toaster but I have repaired one (they're actually very
| electrically and mechanically simple devices). Besides
| the disposable income to build a race car, nothing on
| that list stands out to me as unachievable by anyone who
| chooses to do so.
| hackyhacky wrote:
| > The peer commenter shared the Heinlein quote, but
| there's really something to be said for /society/ of
| being peopled by well-rounded individuals that are able
| to competently turn themselves to many types of tasks
|
| Being a well-rounded individualist is a great, but that's
| an orthogonal issue to the question of outsourcing our
| skills to machinery. When you were growing crops, did you
| till the land by hand or did you use a tractor? When you
| were making clothes did you sew by hand or use a sewing
| machine? Who made your sewing needles?
|
| The (dubious) argument for AI is that using LLMs to write
| code is the same as using modern construction equipment
| to build a house: you get the same result for less
| effort.
| mechagodzilla wrote:
| I've done all of those except tend livestock and build a
| house, but I could probably figure those out with some
| effort.
| kenjackson wrote:
| > I don't know how to grow crops, build a house, tend
| livestock, make clothes, weld metal, build a car, build a
| toaster, design a transistor, make an ASIC, or write an
| OS. I do know how to write a web site. But if I cede that
| skill to an automated process, then that is the feather
| that will break the camel's back?
|
| Reminds me of the Nate Bargatz set where he talks about
| how if he was a time traveler to the past that he
| wouldn't be able to prove it to anyone. The skills most
| of us have require this supply chain and then we apply it
| at the very end. I'm not sure anyone in 1920 cares about
| my binary analysis skills.
| c22 wrote:
| Specialization is over-rated. I've done everything in
| your list except _make an ASIC_ because learning how to
| do those things was interesting and I prefer when things
| are done my way.
|
| I started way back in my 20s _just_ figuring out how to
| write websites. I 'm not sure where the camel's back
| would have broken.
|
| It has, of course, been convenient to be able to
| "bootstrap" my self-reliance in these and other fields by
| consuming goods produced by others, but there is no
| mechanical reason that said goods should be provided by
| specialists rather than advanced generalists beyond our
| irrational social need for maximum acceleration.
| jplusequalt wrote:
| >50 years from new the career of a "coder" will evoke the
| same historical quaintness as does "switchboard operator"
| or "wainwright."
|
| And what happens to those coders? For that matter--what
| happens to all the other jobs at risk of being replaced
| by AI? Where are all the high paying jobs these
| disenfranchised laborers will flock to when their
| previous careers are made obsolete?
|
| We live in a highly specialized society that requires
| people take out large loans to learn the skills necessary
| for their careers. You take away their ability to provide
| their labor, and it now seriously threatens millions of
| workers from obtaining the same quality of life they once
| had.
|
| I seriously oppose such a future, and if that makes me a
| Luddite, so be it.
| hackyhacky wrote:
| > And what happens to those coders? For that matter--what
| happens to all the other jobs at risk of being replaced
| by AI?
|
| Some will manage to remain in their field, most won't.
|
| > Where are all the high paying jobs these
| disenfranchised laborers will flock to when their
| previous careers are made obsolete?
|
| They don't exist. Instead they'll take low-paying jobs
| that can't (yet) be automated. Maybe they'll work in
| factories [1].
|
| > I seriously oppose such a future, and if that makes me
| a Luddite, so be it.
|
| Like I said, the Luddites were right, in the short term.
| In the long term, we don't know. Maybe we'll live in a
| post-scarcity Star Trek world where human labor has been
| completely devalued, or maybe we'll revert to a feudal
| society of property owners and indentured servants.
|
| [1] https://www.newsweek.com/bessent-fired-federal-
| workers-manuf...
| jplusequalt wrote:
| >They don't exist. Instead they'll take low-paying jobs
| that can't (yet) be automated. Maybe they'll work in
| factories
|
| >or maybe we'll revert to a feudal society of property
| owners and indentured servants.
|
| We as the workers in society have the power to see that
| this doesn't happen. We just need to organize. Unionize.
| Boycott. Organize with people in your community to spread
| worker solidarity.
| DrillShopper wrote:
| There is no industry that I have worked in that fights
| against creating or joining unions tooth, claw, and nail
| quite like software engineers.
| jplusequalt wrote:
| I think more and more workers are warming up to unions.
| As wages in software continue to be oppressed, I think
| we'll see an increase in unionization efforts for
| software engineers.
| dTal wrote:
| I dunno, the "tool" that LLMs "replace" is _thinking
| itself_. That seems qualitatively different than anything
| that has come before. It 's the "tool" that underlies
| _all_ the others.
| whilenot-dev wrote:
| > I don't know how to grow crops, build a house, tend
| livestock, make clothes, weld metal, build a car, build a
| toaster, design a transistor, make an ASIC, or write an
| OS. I do know how to write a web site. But if I cede that
| skill to an automated process, then that is the feather
| that will break the camel's back?
|
| All the things you mention have a certain objective
| quality that can be reduced to an approachable minimum. A
| house could be a simple cabin, a tent, a cave; a piece of
| cloth could just be a cape; metal can be screwed, glued
| or cast; a transistor could be a relay or a wooden
| mechanism etc. ...history tells us all that.
|
| I think when there's a _Homo ludens_ that wants to play,
| or when there 's a _Homo economicus_ that wants us to
| optimize, there might be one that separates the process
| of _learning_ from _adaptation_ ( _Homo
| investigans_?)[0]. The process of learning something new
| could be such a subjective property that keeps a yet
| unknown natural threshold which can 't be lowered (or
| "reduced") any further. If I were to be overly
| pessimistic, a hardcore luddite, I'd say that this
| species is under attack, and there will be a generation
| that lacks this aspect, but also won't miss it, because
| this character could have never been experienced in the
| first place.
|
| [0]: https://en.wikipedia.org/wiki/Names_for_the_human_sp
| ecies#Li...
| harrall wrote:
| I don't think specialization is a bad thing but the
| friends I know that only know their subject seem to...
| how do I put this... struggle at life and everything a
| lot more.
|
| And even at work, the coworkers that don't have a lot of
| general knowledge seem to work a lot harder and get less
| done because it takes them so much longer to figure
| things out.
|
| So I don't know... is avoiding the work of learning worth
| it to struggle at life more?
| bko wrote:
| I don't know, most of the things I'm reliant on, from my
| phone, ISP, automobile, etc are built on fragile
| interdependent supply chains provided by for-profit
| companies. If you're really worried about this, you should
| learn survival skills not the academic topics I'm talking
| about.
|
| So if you're not bothering to learn how to farm, dress some
| wild game, etc, chances are this argument won't be
| convincing for "why should I learn calculus"
| Zambyte wrote:
| For what it's worth, locally runnable language models are
| becoming exceptionally capable these days, so if you assume
| you will have some computer to do computing, it seems
| reasonable to assume that it will enable you to do some
| language model based things. I have a server with a single
| GPU running language models that easily blow GPT 3.5 out of
| the water. At that point, I am offloading reasoning tasks
| to my computer in the same way that I offload memory take
| to my computer through my note taking habits.
| whatshisface wrote:
| I don't think memorizing poetry fits your picture. Nobody
| ever memorized poetry so that they could answer questions
| about it.
| bko wrote:
| A large part was to preserve cultural knowledge, which is
| kind of like answering questions about it. What wisdom or
| knowledge does this entail. People do the same with
| religious texts today
|
| The other part I imagine was largely entertainment, social
| and memory is a good skill to build.
| KronisLV wrote:
| It doesn't seem that different from having to write a book
| report or something like that. Back in school, we also
| needed to memorize poems and songs to recite them - I quite
| hated it because my memory was never exactly great. Same as
| having to remember the vocabulary in a foreign language
| when learning it, though that might arguably be a bit more
| directly useful.
| AlexandrB wrote:
| > For instance, I'm not too concerned about my child's
| ability to write very legibly (most writing is done on
| computers), spell very well (spell check keeps us
| professional), reading a map to get around (GPS), etc
|
| What I don't like are all the hidden variables in these
| systems. Even GPS, for example, is making some assumptions
| about what kind of roads you want to take and how to weigh
| different paths. LLMs are worse in this regard because the
| creators encode a set of moral and stylistic
| assumptions/dictates into the model and everybody who uses it
| is nudged into that paradigm. This is destructive to any kind
| of original thought, especially in an environment where there
| are only a handful of large companies providing the models
| everyone uses.
| Retric wrote:
| The scope of what's useful to know changes with tools, but
| having a bullshit detector requires actually knowing some
| things and being able to reason about the basics.
|
| It's not that LLM's are particularly different it's that
| people are less able to determine when they are messing up. A
| search engine fails and you notice, an LLM fails and your
| boss, customer, ect notices.
| andai wrote:
| >Why should I learn to do X if I can just ask an LLM and it
| will do it better than me
|
| This may eventually apply to all human labor.
|
| I was thinking, even if they pass laws to mandate companies
| employ a certain fraction of human workers... it'll be like
| it already is now: they just let AI do most of the work
| anyway!
| riohumanbean wrote:
| Why have children learn to walk? They're better off learning
| the newest technology of hoverboards and not getting left
| behind!
| CivBase wrote:
| Why should you learn how to add when you can just use a
| calculator? We've had calculators for decades!
|
| Because understanding how addition works is instrumental to
| understanding more advanced math concepts. And being able to
| perform simple addition quickly, without a calculator is a
| huge productivity boost for many tasks.
|
| In the world of education and intellectual development it's
| not about getting the right answer as quickly as possible.
| It's about mastering simple things so that you can understand
| complicated things. And often times mastering a simple thing
| requires you to manually do things which technology could
| automate.
| light_hue_1 wrote:
| > For instance, I'm not too concerned about my child's
| ability to write very legibly (most writing is done on
| computers), spell very well (spell check keeps us
| professional), reading a map to get around (GPS), etc.
|
| I'm the polar opposite. And I'm AI researcher.
|
| The reason you can't answer your kid when he asks about LLMs
| is because the original position was wrong.
|
| Being able to write isn't optional. It's a critical tool for
| thought. Spelling is very important because you need to avoid
| confusion. If you can't spell no spell checker can save you
| when it inserts the wrong word. And this only gets far worse
| the more technical the language is. And maps are crucial too.
| Sometimes, the best way to communicate is to draw a map. In
| many domains like aviation maps are everything, you literally
| cannot progress without them.
|
| LLMs are no different. They can do a little bit of thinking
| for us and help us along the way. But we need to understand
| what's going on to ask the right questions and to understand
| their answers.
| jplusequalt wrote:
| >children will have a different perspective
|
| Children will lack the critical thinking for solving complex
| problems, and even worse, won't have the work ethic for
| dealing with the kinds of protracted problems that occur in
| the real world.
|
| But maybe that's by design. I think the ownership class has
| decided productivity is more important than societal malaise.
| bcrosby95 wrote:
| > So how do I answer my child when he asks "Why should I
| learn to do X if I can just ask an LLM and it will do it
| better than me"
|
| It's been my experience that LLMs are only better than me at
| stuff I'm bad at. It's noticeably worse than me at things I'm
| good at. So the answer to your question depends: can your
| child get good at things while leaning on an LLM?
|
| I don't know the answer to this. Maybe schools need to expect
| more from their students with LLMs in the picture.
| gnatolf wrote:
| Given the rate of improvement wrt to llms, this may not
| hold true for long
| johndough wrote:
| Use it or lose it. With the invention of the calculator,
| students lost the ability to do arithmetic. Now, with LLMs,
| they lose the ability to think.
|
| This is not conjecture by the way. As a TA, I have observed
| that half of the undergraduate students lost the ability to
| write any code at all without the assistance of LLMs. Almost
| all use ChatGPT for most exercises.
|
| Thankfully, cheating technology is advancing at a similarly
| rapid pace. Glasses with integrated cameras, WiFi and heads-
| up display, smartwatches with polarized displays that are
| only readable with corresponding glasses, and invisibly small
| wireless ear-canal earpieces to name just a few pieces of
| tech that we could have only dreamed about back then. In the
| end, the students stay dumb, but the graduation rate barely
| suffers.
|
| I wonder whether pre-2022 degrees will become the academic
| equivalent to low-background radiation steel:
| https://en.wikipedia.org/wiki/Low-background_steel
| quantumHazer wrote:
| Universities still teach you calculus and real analysis even
| though Wolfram Alpha exists. It boils down to your willing to
| learn something. An LLM can't understand things for you. I'm
| "early genz" and I write code without llm because I find data
| structure and algorithm very interesting and I want to learn
| the concepts not because I'm in love with the syntax of C or
| Rust (I love the syntax of C btw).
| an_aparallel wrote:
| This is how we end up with people who cant write legibly,
| cant smell bad maths (on the news/articles/ads), cant change
| tires, have no orienteering or sense of direction and
| memories like swiss cheese. Trust the oracle son. /s
|
| I think all of the above do one thing brilliantly, built self
| confidence.
|
| Its easy to get bullshitted if what youre able to hold in
| your head is effectively nothing.
| palmotea wrote:
| > When modern search became more available, a lot of people
| said there's no point of rote memorization as you can just do
| a Google search. That's more or less accepted today.
|
| And those people are wrong, in a similar way to how it's
| wrong to say: "There's no point in having very much RAM, as
| you can just page to disk."
|
| It's the cognitive equivalent of becoming morbidly obese
| (another popular decision in today's world).
| OptionOfT wrote:
| The problem with GPS is that you never learn to orient
| yourself. You don't learn to have a sense of place, direction
| or elapsed distance. [0]
|
| As to writing, just the action of writing something down with
| a pen, on paper, has been proven to be better for
| memorization than recording it on a computer [1].
|
| If we're not teaching these basic skills because an LLM does
| it better, how do learn to be skeptical of the output of the
| LLM. How do we validate it?
|
| How do we bolster ourselves against corporate influences when
| asking which of 2 products is healthier? How do we spot
| native advertising? [2]
|
| [0]: https://www.nature.com/articles/531573a
|
| [1]: https://www.sciencedirect.com/science/article/abs/pii/S0
| 0016...
|
| [2]: Example: https://www.nytimes.com/paidpost/netflix/women-
| inmates-separ...
| taftster wrote:
| I don't think asking "what's wrong with my code" hurts the
| learning process. In fact, I would argue it helps it. I don't
| think you learn when you have reached your frustration point
| and you just want the dang assignment completed. But before
| reaching that point, if you had a tutor or assistant that you
| could ask, "hey, I'm just not seeing my mistake, do you have
| ideas" goes a long way to foster learning. ChatGPT, used in
| this way, can be extremely valuable and can definitely unlock
| learning in new ways which we probably even haven't seen yet.
|
| That being said, I agree with you, if you just ask ChatGPT to
| write a b-tree implementation from scratch, then you have not
| learned anything. So like all things in academia, AI can be
| used to foster education or cheat around it. There's been
| examples of these "cheats" far before ChatGPT or Google
| existed.
| SoftTalker wrote:
| No I think the struggle is essential. If you can just ask a
| tutor (real or electronic) what is wrong with your code, you
| stop thinking and become dependent on that. Learning to think
| your way through a roadblock that seems like a showstopper is
| huge.
|
| It's sort of the mental analog of weight training. The only
| way to get better at weightlifting is to actually lift
| weight.
| taftster wrote:
| If I were to go and try to bench 300lbs, I would absolutely
| need a spotter to rescue me. Taking on more weight than I
| can possibly achieve is a setup for failure.
|
| Sure, I should probably practice benching 150lbs. That
| would be a good challenge for me and I would benefit from
| that experience. But 300lbs would crush me.
| Maskawanian wrote:
| Agreed, the only thing that is certain is that they are
| cheating themselves.
|
| While it can be useful to use LLMs as a tutor if you're stuck.
| The moment that you use it to provide a solution, you stop
| learning and the tool becomes a required stepping stone.
| hobo_in_library wrote:
| The challenge is that while LLMs do not know everything, they
| are likely to know everything that's needed for your
| undergraduate education.
|
| So if you use them at that level you may learn the concepts at
| hand, but you won't learn _how to struggle_ to come up with
| novel answers. Then later in life when you actually hit problem
| domains that the LLM wasn't trained in, you'll not have learned
| the thinking patterns needed to persist and solve those
| problems.
|
| Is that necessarily a bad thing? It's mixed: - You lower the
| bar for entry for a certain class of roles, making labor
| cheaper and problems easier to solve at that level. - For more
| senior roles that are intrinsically solving problems without
| answers written in a book or a blog post somewhere, you need to
| be selective about how you evaluate the people who are ready to
| take on that role.
|
| It's like taking the college weed out classes and shifting
| those to people in the middle of their career.
|
| Individuals who can't make the cut will find themselves
| stagnating in their roles (but it'll also be easier for them to
| switch fields). Those who can meet the bar might struggle but
| can do well.
|
| Business will also have to come up with better ways to evaluate
| candidates. A resume that says "Graduated with a degree in X"
| will provide less of a signal than it did in the past
| psygn89 wrote:
| Agreed, the struggle often leads us to poke and prod an issue
| from many angles until things finally click. It lets us think
| critically. In that journey you might've learned other related
| concepts which further solidifies your understanding.
|
| But when the answer flows out of thin air right in front of you
| with AI, you get the "oh duh" or "that makes sense" moments and
| not the "a-ha" moment that ultimately sticks with you.
|
| Now does everything need an "a-ha" moment? No.
|
| However, I think core concepts and fundamentals need those
| "a-ha" moments to build a solid and in-depth foundation of
| understanding to build upon.
| taftster wrote:
| Absolutely this. AI can help reveal solutions that weren't
| seen. An a-ha moment can be as instrumental to learning as
| the struggle that came before.
|
| Academia needs to embrace this concept and not try to fight
| it. AI is here, it's real, it's going to be used. Let's teach
| our students how to benefit from its (ethical) use.
| dyauspitr wrote:
| I'm pretty sure you can assume close to 100% of students are
| using LLMs to do their homework.
| ryandrake wrote:
| And if you're that one person out of 100,000 who is _not_
| using LLMs to do their homework, you are at a significant
| disadvantage on the grading curve.
| DontchaKnowit wrote:
| Maybe, but piss on that, who needs good grades? Youll learn
| a hell of a lot better
| andai wrote:
| I spent much of the past year at public libraries, and I heard
| the word ChatGPT approximately once per minute, in surround
| sound. Always from young people, and usually in a hushed
| tone...
| moltar wrote:
| I think it's finally time to just stop the homework.
|
| All school work must be done within the walls of the school.
|
| What are we teaching our children? It's ok to do more work at
| home?
|
| There are countries that have no homework and they do just
| fine.
| jplusequalt wrote:
| Homework helps reinforce the material learned in class. It's
| already a problem where there is too much material to be fit
| into a single class period. Trying to cram in enough time for
| homework will only make that problem worse.
| moltar wrote:
| Can do the work the next day to reinforce.
|
| As I said there are countries without homework and they
| seem to do ok. So it's not mandatory by any means.
| jplusequalt wrote:
| >Can do the work the next day to reinforce.
|
| Keeping the curriculum fixed, there's already barely
| enough time to cover everything. Cutting the amount of
| lectures in half to make room for in-class homework time
| does not fix this fundamental problem.
| DontchaKnowit wrote:
| Just make lecture times longer.
| srveale wrote:
| IMO it's so easy to ChatGPT your homework that the whole
| education model needs to flip on its head. Some teachers
| already do something like this, it's called the "Flipped
| classroom" approach.
|
| Basically, a student's marks depend mostly (only?) on what they
| can do in a setting where AI is verifiably unavailable. It
| means less class time for instruction, but students have a
| tutor in their pocket anyway.
|
| I've also talked with a bunch of teachers and a couple admins
| about this. They agree it's a huge problem. By the same token,
| they are using AI to create their lesson plans and assignments!
| Not fully of course, they edit the output using their
| expertise. But it's funny to imagine AI completing an AI
| assignment with the humans just along for the ride.
|
| The point is, if you actually want to know what a student is
| capable of, you need to watch them doing it. Assigning homework
| has lost all meaning.
| hackyhacky wrote:
| > it's called the "Flipped classroom" approach.
|
| Flipped classroom is just having the students give lectures,
| instead of the teacher.
|
| > Basically, a student's marks depend mostly (only?) on what
| they can do in a setting where AI is verifiably unavailable.
|
| This is called "proctored exams" and it's been pretty common
| in universities for a few centuries.
|
| None of this addresses the real issue, which is whether
| teachers _should_ be preventing students from using AIs.
| srveale wrote:
| > Flipped classroom is just having the students give
| lectures, instead of the teacher.
|
| Not quite. Flipped classroom means more instruction outside
| of class time and less homework.
|
| > This is called "proctored exams" and it's been pretty
| common in universities for a few centuries. None of this
| addresses the real issue
|
| Proctored exams is part of it. In-class assignments is
| another. Asynchronous instruction is another.
|
| And yes, it addresses the issue. Students can use AI
| however they see fit, to learn or to accomplish tasks or
| whatever, but for actual assessment of ability they cannot
| use AI. And it leaves the door open for "open-book" exams
| where the use of AI is allowed, just like a calculator and
| textbook/cheat-sheet is allowed for some exams.
|
| https://en.wikipedia.org/wiki/Flipped_classroom
| bryanlarsen wrote:
| Flipped classroom means you watch the recorded lecture
| outside of class time and you do your homework during class
| time.
| Spivak wrote:
| Thank you, it's amazing how people don't even try to
| understand what words mean before dismissing it. Flipped
| makes way more sense anyway since lectures aren't
| terribly interactive. Being able to pause/replay/skip
| around in lectures is underrated.
| vonneumannstan wrote:
| >Not fully of course, they edit the output using their
| expertise
|
| Surely this is sarcasm, but really your average schoolteacher
| is now a C student Education Major.
| srveale wrote:
| I was talking about people I know and talk with, mostly
| friends and family, who are smart, hard working, and their
| students are lucky to have them.
| chalst wrote:
| Students who do that risk submitting assignments that show they
| don't understand the course so far.
| 0xffff2 wrote:
| >For many students, it's literally "let me paste the assignment
| into ChatGPT and see what it spits out, change a few words and
| submit that".
|
| Does that actually work? I'm long past having easy access to
| college programming assignments, but based on my limited
| interaction with ChatGPT I would be absolutely shocked if it
| produced output that was even coherent, much less working code
| given such an approach.
| StefanBatory wrote:
| I have some subjects, at Masters - that are solvable by one
| prompt. One.
|
| Quality of CS/Software Engineering programs vary that much.
| bongodongobob wrote:
| Why are you asking? Go try it. And yes, depending on the
| task, it does.
| 0xffff2 wrote:
| As I said, I'm not a student, so I don't have access to a
| homework assignment to paste in. Ironically I have pretty
| much everything I ever submitted for my undergrad, but it
| seems like I absolutely never archived the assignments for
| some reason.
| rufus_foreman wrote:
| >> Does that actually work?
|
| Sure. Works in my IDE. "Create a linked list implementation,
| use that implementation in a method to reverse a linked list
| and write example code to demonstrate usage".
|
| Working code in a few seconds.
|
| I'm very glad I didn't have access to anything like that when
| I was doing my CS degree.
| ryandrake wrote:
| I think teachers also need to reconsider how they are measuring
| mastery in the subject. LLMs exist. There is no putting the cat
| back into the bag. If your 1980s way to measure a student's
| mastery of a subject can be fooled by an LLM, then how
| effective is that measurement in 2020+? Maybe we need to stop
| using essays as a way to tell if the student has learned the
| material.
|
| Don't ask me what the solution is. Maybe your product does it.
| If I knew, I'd be making a fortune selling it to universities.
| teekert wrote:
| Students do something akin to vibe coding I guess. It may seem
| impressive at first glance but if anything breaks you are so,
| so lost. Maybe that's it, break the student's code, see how
| they fix it. The vibe coding student is easily separate from
| the real one (of course this real coder can also use AI, just
| not yoloing it).
|
| I guess you can apply similar mechanics to reports. Some deeper
| questions and you will know if the report was self written or
| if an AI did it.
| bboygravity wrote:
| I don't get this reasoning. Without LLMs I would learn how to
| write sub-optimal code that is somewhat functional. With LLMs
| instantly see "how it's done" for my exact problem case which
| makes me learn way faster. On top of that it always makes dumb
| mistakes which forces you to actually understand what it's
| spitting out to get it to work properly. Again: that helps with
| learning.
|
| The fact that you can ask it for a solution for exactly the
| context you're interested in is amazing and traditional
| learning doesn't come close in terms of efficiency IMO.
| dingnuts wrote:
| > With LLMs instantly see "how it's done" for my exact
| problem case which makes me learn way faster.
|
| No, you see a plausible set of tokens that appear similar to
| how it's done, and as a beginner, you're not able to tell the
| difference between a good example and something that is
| subtly wrong.
|
| So you learn something, but it's wrong. You internalize it.
| Later, it comes back to bite you. But OpenAI keeps the money
| for the tokens. You pay whether the LLM is right or not. Sam
| likes that.
| Spivak wrote:
| This makes for a good sound bite but it's just not true.
| The use case of "show me what is a customary solution to
| <problem>" plays exactly into LLMs strength as a funny kind
| of search engine. I used to (and still do) search public
| code for this use case to get a sense of the style and
| idioms common in a new language/library and the plausible
| set of tokens is doing exactly that.
| victorbjorklund wrote:
| In one way I'm glad I learned to code before LLM:s. It would be
| so hard to push through the learning now when you are just a
| click away from buildning the app with AI...
| dmurray wrote:
| I am surprised that business students are relatively low
| adopters: LLMs seem perfect for helping with presentations, etc,
| and business students are stereotypically practical-minded rather
| than motivated by love of the subject.
|
| Perhaps Claude is disproportionately _marketed_ to the STEM
| crowd, and the business students are doing the same stuff using
| ChatGPT.
| jimbob45 wrote:
| It says STEM undergrad students are the primary beneficiaries of
| LLMs but Wolfram Alpha was already able to do the lion's share of
| most undergrad STEM homework 15 years ago.
| brunocroh wrote:
| I simply don't waste my time reading an AD as an article.
|
| I take this as seriously as I would if McDonald's published
| articles about how much weight people lose eating at McDonald's.
| ikesau wrote:
| It's more like an analysis of what items people order from
| McDonald's, using McDonald's own data which is otherwise very
| difficult to collect.
|
| Your loss!
| brunocroh wrote:
| Yes, maybe, but there is a lot of noise and conflicts of
| interest.
| AlexandrB wrote:
| This is why I go to cigarette companies for analysis of the
| impact of smoking on users. They have the most data!
| lblume wrote:
| If you had read the article, you would have been able to see
| that the conclusions don't really align with any economic goals
| Anthropic might have.
| AlexandrB wrote:
| I think the point is that the situation is probably worse
| than what Anthropic is presenting here. So if the conclusions
| are just damaging, the reality must be truly damning.
| defgeneric wrote:
| To have the reputation as an AI company that really cares
| about education and the responsible integration of AI into
| education is a pretty valuable goal. They are now ahead of
| OpenAI in this respect.
|
| The problem is that there's a conflict of interest here. The
| extreme case proves it--leaving aside the feasibility of it,
| what if the only solution is a total ban on AI usage in
| education? Anthropic could never sanction that.
| juped wrote:
| I'm curious if you're willing to say what you (and potentially
| other people who spell 'AD' like that) think it's an acronym
| for, by the way.
| karpour wrote:
| My take: While AI tools can help with learning, the vast majority
| of students use it to _avoid_ learning
| hervature wrote:
| This has been observation about the internet. Growing up in a
| small town without access to advanced classes, having access to
| Wikipedia felt like the greatest equalizer in the world. 20
| years post internet, seeing the most common outcome be that
| people learn less as a result of unlimited access to
| information would be depressing if it did not result in my own
| personal gain.
| karpour wrote:
| I would say a big difference of the Internet around 2000 and
| the internet now is that most people shared information in
| good faith back then, which is not the case anymore. Maybe
| back then people were just as uncritical of information, but
| now we really see the impact of people being not critical.
| nthingtohide wrote:
| My take : AI is the REPL interface for learning activities. All
| the points which Salman Khan talked about apply here.
| janalsncm wrote:
| I agree with you, but I hope schools also take the opportunity
| to reflect on what they teach and how. I used to think I hated
| writing, but it turns out I just hated English class. (I got a
| STEM degree because I hated English class so much, so maybe I
| have my high school English teacher to thank for it.)
|
| Torturing students with five paragraph essays, which is what
| "learning" looks like for most American kids, is not that great
| and isn't actually teaching critical thinking which is most
| valuable. I don't know any other form of writing that is like
| that.
|
| Reading "themes" into books that your teacher is convinced are
| there. Looking for 3 quotes to support your thesis (which must
| come in the intro paragraph, but not before the "hook" which
| must be exciting and grab the reader's attention!).
| ozarkerD wrote:
| I loved asking questions as a kid. To the point of annoying
| adults. I would have loved to sit and ask these AI questions
| about all kinds of interests when I was young.
| qwertox wrote:
| I'm pretty sure that kids at the age of 4 would get an amazing
| intelligence boost compared to their peers later when they are
| around 8 years old.
|
| They will clearly recognize other kids which did not have an AI
| to talk with at that stage when curiosity really blossoms.
| walthamstow wrote:
| I think it's likely that everyone here was, or even is, that
| kid and that's why we're here on this website today
| PeterStuer wrote:
| I feel CS students, and to a lesser degree STEM in general, will
| always be more early adopters of advancements in computer
| technology.
|
| They were the first to adopt digital wordprocessing,
| presentations, printing and now generative AI even though in
| essence all of these would have been disproportionately more hand
| in glove for the humanities on a purely functional level.
|
| It's just a matter of comfortability with and interest in
| technology.
| chenzo44 wrote:
| professor here. i set up a website to host openwebui to use in my
| b-school courses (UG and grad). the only way i've found to get
| students to stop using it to cheat is to push them to use it
| until they learn for themselves that it doesn't answer everything
| correctly. this requires careful thoughtful assignment redesign.
| everytime i grade a submission with the hallmarks of ai-
| generation, i always find that it fails to cite content from the
| course and shows a lack of depth. so, i give them the grade they
| earn. so much hand wringing about using ai to cheat... just
| uphold the standards. if they are so low that ai can easily game
| them, that's on the instructor.
| fudged71 wrote:
| an interesting area potentially missed (though acknowledged as
| out of scope) is how students might use LLMs for tasks related to
| early adulthood development. Successfully navigating post-
| secondary education involves more than academics; it requires
| developing crucial life skills like resilience, independence,
| social integration, and well-being management, all of which are
| foundational to academic persistence and success. Understanding
| if and how students leverage AI for these non-academic,
| developmental challenges could offer a more holistic picture of
| AI's role in student life and its indirect impact on their
| educational journey
| zebomon wrote:
| The writing is irrelevant. Who cares if students don't learn how
| to do it? Or if the magazines are all mostly generated a decade
| from now? All of that labor spent on writing wasn't really making
| economic sense.
|
| The problem with that take is this: it was never _about_ the act
| of writing. What we lose, if we cut humans out of the equation,
| is writing as a proxy for what actually matters, which is
| thinking.
|
| You'll soon notice the downsides of not-thinking (at scale!) if
| you have a generation of students who weren't taught to exercise
| their thinking by writing.
|
| I hope that more people come around to this way of seeing things.
| It seems like a problem that will be much easier to mitigate than
| to fix after the fact.
|
| A little self-promo: I'm building a tool to help students and
| writers create proof that they have written something the good ol
| fashioned way. Check it out at https://itypedmypaper.com and let
| me know what you think!
| spongebobstoes wrote:
| Writing is not necessary for thinking. You can learn to think
| without writing. I've never had a brilliant thought while
| writing.
|
| In fact, I've done a lot more thinking and had a lot more
| insights from talking than from writing.
|
| Writing can be a useful tool to help with rigorous thinking. In
| my opinion, is mostly about augmenting the author's effective
| memory to be larger and more precise.
|
| I'm sure the same effect could be achieved by having AI
| transcribe a conversation.
| Unearned5161 wrote:
| I'm not settled on transcribed conversation being an adequate
| substitute for writing, but maybe it's better than nothing.
|
| There's something irreplaceable about the absoluteness of
| words on paper and the decisions one has to do to write them
| out. Conversational speak is, almost by definition, more
| relaxed and casual. The bar is lower and as such, the bar for
| thoughts is lower, in order of ease of handwaving I think it
| goes: mental, speech, writing.
|
| Furthermore there's the concept of editing which I'm unsure
| how it could be carried out in a conversational sense in
| graceful manner. Being able to revise words, delete, move
| around, can't be done with conversation unless you count
| "forget I said that, it's actually more like this..." as
| suitable.
| janalsncm wrote:
| How does your product prevent a person from simply retyping
| something that ChatGPT wrote?
|
| I think the prevalence of these AI writing bots means schools
| will have to start doing things that aren't scalable: in-class
| discussions, in-person writing (with pen and paper or locked
| down computers), way less weight given to remote assignments on
| Canvas or other software. Attributing authorship from text
| alone (or keystroke patterns) is not possible.
| zebomon wrote:
| It may be possible that with enough data from the two
| categories (copied from ChatGPT and not), your keystroke
| dynamics will differ. This is an open question that my co-
| founder and I are running experiments on currently.
|
| So, I would say that while I wouldn't fully dispute your
| claim that attributing authorship from text alone is
| impossible, it isn't yet totally clear one way or the other
| (to us, at least -- would welcome any outside research).
|
| Long-term -- and that's long-term in AI years ;) -- gaze
| tracking and other biometric tracking will undoubtedly be
| necessary. At some point in the near future, many people will
| be wearing agents inside earbuds that are not obvious to the
| people around them. That will add another layer of complexity
| that we're aware of. Fundamentally, it's more about creating
| evidence than creating proof.
|
| We want to give writers and students the means to create
| something more detailed than they would get from a chatbot
| out-of-the-box, so that mimicking the whole act of writing
| becomes more complicated.
| pr337h4m wrote:
| At this point, it would be easier to stick to in-person
| assignments.
| zebomon wrote:
| It certainly would be! I think for many students though,
| there's something lost there. I was a student who got a
| lot more value out of my take-home work than I did out of
| my in-class work. I don't think that I ever would have
| taken the interest in writing that I did if it wasn't
| such a solitary, meditative thing for me.
| logicchains wrote:
| >I think the prevalence of these AI writing bots means
| schools will have to start doing things that aren't scalable
|
| It won't be long 'til we're at the point that embodied AI can
| be used for scalable face-to-face assessment that can't be
| cheated any easier than a human assessor.
| ketzu wrote:
| > The writing is irrelevant.
|
| In my opinion this is not true. Writing is a form of
| communicating ideas. Structuring and communicating ideas with
| others is really important, not just in written contexts, and
| it needs to be trained.
|
| Maybe the way universities do it is not great, but writing in
| itself is important.
| zebomon wrote:
| Kindly read past the first line, friend :)
| knowaveragejoe wrote:
| Paul Graham had a recent blogpost about this, and I find it
| hard to disagree with.
|
| https://www.paulgraham.com/writes.html
| karn97 wrote:
| I literally never write while thinking lol stop projecting this
| hard
| technoabsurdist wrote:
| I'm an undergrad at a T10 college. Walking through our library, I
| often notice about 30% of students have ChatGPT or Claude open on
| their screens.
|
| In my circle, I can't name a single person who doesn't heavily
| use these tools for assignments.
|
| What's fascinating, though, is that the most cracked CS students
| I know deliberately avoid using these tools for programming work.
| They understand the value in the struggle of solving technical
| problems themselves. Another interesting effect: many of these
| same students admit they now have more time for programming and
| learning they "care about" because they've automated their
| humanities, social sciences, and other major requirements using
| LLMs. They don't care enough about those non-major courses to
| worry about the learning they're sacrificing.
| proteal wrote:
| I'm about to graduate from a top business school with my MBA and
| it's been wild seeing AI evolve over the last 2 years.
|
| GPT3 was pretty ass - yet some students would look you dead in
| the eyes with that slop and claim it as their own. Fast forward
| to last year when I complimented a student on his writing and he
| had to stop me - "bro this is all just AI."
|
| I've used AI to help build out frameworks for essays and suggest
| possible topics and it's been quite helpful. I prefer to do the
| writing myself because the AIs tend to take very bland positions.
| The AIs are also great at helping me flesh out my writing. I ask
| "does this make sense" and it tells me patiently where my writing
| falls off the wagon.
|
| AI is a game changer in a big way. Total paradigm shift. It can
| now take you 90% of the way with 10% of the effort. Whether this
| is good or bad is beyond my pay grade. What I can say is that if
| you are not leveraging AI, you will fall behind those that are.
| moojacob wrote:
| How can I, as a student, avoid hindering my learning with
| language models?
|
| I use Claude, a lot. I'll upload the slides and ask questions.
| I've talked to Claude for hours trying to break down a problem. I
| think I'm learning more. But what I think might not be what's
| happening.
|
| In one of my machine learning classes, cheating is a huge issue.
| People are using LMs to answer multiple choice questions on
| quizzes that are on the computer. The professors somehow found
| out students would close their laptops without submitting, go out
| into the hallway, and use a LM on their phone to answer the
| questions. I've been doing worse in the class and chalked it up
| to it being grad level, but now I think it's the cheating.
|
| I would never do cheat like that, but when I'm stuck and use
| Claude for a hint on the HW am I loosing neurons? The other day I
| used Claude to check my work on a graded HW question (breaking
| down a binary packet) and it caught an error. I did it on my own
| before and developed some intuition but would I have learned more
| if I submitted that and felt the pain of losing points?
| lunarboy wrote:
| This sounds fine? Copy pasting LLM output without understanding
| is a short term dopamine hit that only hurts you long term if
| you don't understand it. If you struggle first, or
| strategically ping-pong with the LLM to arrive at the answer,
| and can ultimately understand the underlying reasoning.. why
| not use it?
|
| Of course the problem is the much lower barrier for that to
| turn into cutting corners or full on cheating, but always
| remember it ultimately hurts you the most long term.
| azemetre wrote:
| Can you do all this without relying on any LLM usage? If so
| then you're fine.
| knowaveragejoe wrote:
| It's a hard question to answer and one I've been mindful of in
| using LLMs as tutoring aids for my own learning purposes. Like
| everything else around LLM usage, it probably comes down to
| careful prompting... I really _don 't_ want the answer right
| away. I want to propose my own thoughts and carefully break
| them down with the LLM. Claude is pretty good at this.
|
| "productive struggle" is essential, I think, and it's hard to
| tease that out of models that are designed to be as immediately
| helpful as possible.
| dwaltrip wrote:
| Only use LLMs for half of your work, at most. This will ensure
| you continue to solidify your fundamentals. It will also
| provide an ongoing reality check.
|
| I'd also have sessions / days where I don't use AI at all.
|
| Use it or lose it. Your brain, your ability to persevere
| through hard problems, and so on.
| quantumHazer wrote:
| As a student, I use LLMs as little as possible and try to rely
| on books whenever possible. I sometimes ask LLMs questions
| about things that don't click, and I fact-check their
| responses. For coding, I'm doing the same. I'm just raw dogging
| the code like a caveman because I have no corporate deadlines,
| and I can code whatever I want. Sometimes I get stuck on
| something and ask an LLM for help, always using the web
| interface rather than IDEs like Cursor or Windsurf.
| Occasionally, I let the LLMs write some boilerplate for boring
| things, but it's really rare and I tend not to use them too
| much. This isn't due to Luddism but because I want to learn,
| and I don't want slop in my way.
| defgeneric wrote:
| After reading the whole article I still came away with the
| suspicion that this is a PR piece that is designed to head-off
| strict controls on LLM usage in education. There is a fundamental
| problem here beyond cheating (which is mentioned, to their
| credit, albeit little discussed). Some academic topics are only
| learned through sustained, even painful, sessions where attention
| has to be fully devoted, where the feeling of being "stuck" has
| to be endured, and where the brain is given space and time to do
| the real work of synthesizing, abstracting, and learning, or, in
| short, _thinking_. The prompt-chains where students are asking
| "show your work" and "explain" can be interpreted as the kind of
| back-and-forth that you'd hear between a student and a teacher,
| but they could also just be evidence of higher forms of
| "cheating". If students are not really working through the
| exercises at the end of each chapter, but instead offloading the
| task to an LLM, then we're going to have a serious competency
| issue. Nobody ever actually learns anything.
|
| Even in self-study, where the solutions are at the back of the
| text, we've probably all had the temptation to give up and just
| flip to the answer. Anthropic would be more responsible to admit
| that the solution manual to every text ever made is now instantly
| and freely available. This has to fundamentally change pedagogy.
| No discipline is safe, not even those like music where you might
| think the end performance is the main thing (imagine a promising,
| even great, performer who cheats themselves in the education
| process by offloading any difficult work in their music theory
| class to an AI, coming away learning essentially nothing).
|
| P.S. There is also the issue of grading on a curve in the current
| "interim" period where this is all new. Assume a lazy professor,
| or one refusing to adopt any new kind of teaching/grading method:
| the "honest" students have no incentive to do it the hard way
| when half the class is going to cheat.
| juancroldan wrote:
| With so much collaborative usage, I wonder how Claude group chats
| are not already a feature
| atoav wrote:
| As someone teaching at the university level, the goals of
| teaching are (in that order):
|
| 1. Get people interested in my topics and removing fears and/or
| preconceived notions about whether it is something for them or
| not
|
| 2. Teach students general principles and the ability to go deeper
| themselves when and if it is needed
|
| 3. Giving them the ability to apply the learned
| principles/material in situations they encounter
|
| I think removing fear and sparking interest is a precondition for
| the other two. And if people are interested _they_ want to
| understand it and then they use AI to answer questions they have
| instead of blindly letting it do the work.
|
| And even before AI you would have students who thought they did
| themselves favours by going a learn-and-forget route or cheating.
| AI jusr makes it a little easier to do just that. But in any
| pressure situation, like a written assignment under supervision
| it will come to light anyways, whether someone knows their shit
| or not.
|
| Now I have the luck that the topics I teach (electronics and
| media technology) are very applied anyways, so AI does not have a
| big impact as of now. Not being able to understand things isn't
| really an option when you have to use a mixing desk in a venue
| with a hundred people or when you have to set up a tripod without
| wrecking the 6000EUR camera on top.
|
| But I generally teach people who are in it for the interest and
| not for some prestige that comes with having a BA/MA. I can
| imagine this is quite different in other fields where people are
| in it for the money or the prestige.
| mceoin wrote:
| I'm curious why people think business is so underrepresented as a
| user group, especially since "analyzing" 30% of the Bloom
| Taxonomy results. My dual theories are:
|
| - LLMs are good enough to zero or few-shot most business
| questions and assignments, so n.questions is low VS other tasks
| like writing a codebase.
|
| - Form factor (biased here); maybe threads-only aren't best for
| business analysis?
| bsoles wrote:
| My BS detector went up to 11 as I was reading the article. Then I
| realized that "Education Report" was written by Anthropic itself.
| The article is a prime example of AI-washing.
|
| > Students primarily use AI systems for creating...
|
| > Direct conversations, where the user is looking to resolve
| their query as quickly as possible
|
| Aka cheating.
| xcke wrote:
| This topic is also interesting to me because I have small
| children.
|
| Currently, I view LLMs as huge enablers. They helped me create a
| side-project alongside my primary job, and they make development
| and almost anything related to knowledge work more interesting. I
| don't think they made me think less; rather, they made me think a
| lot more, work more, and absorb significantly more information.
| But I am a senior, motivated, curious, and skilled engineer with
| 15+ years of IT, Enterprise Networking, and Development
| experience.
|
| There are a number of ways one can use this technology. You can
| use it as an enabler, or you can use it for cheating. The
| education system needs to adapt rapidly to address the challenges
| that are coming, which is often a significant issue (particularly
| in countries like Hungary). For example, consider an exam where
| you are allowed to use AI (similar to open-book exams), but the
| exam is designed in such a way that it is sufficiently difficult,
| so you can only solve it (even with AI assistance) if you possess
| deep and broad knowledge of the domain or topic. This is doable.
| Maybe the scoring system will be different, focusing not just on
| whether the solution works, but also on how elegant it is. Or, in
| the Creator domain, perhaps the focus will be on whether the
| output is sufficiently personal, stylish, or unique.
|
| I tend to think current LLMs are more like tools and enablers. I
| believe that every area of the world will now experience a boom
| effect and accelerate exponentially.
|
| When superintelligence arrives--and let's say it isn't sentient
| but just an expert system--humans will still need to chart the
| path forward and hopefully control it in such a way that it
| remains a tool, much like current LLMs.
|
| So yes, education, broad knowledge, and experience are very
| important. We must teach our children to use this technology
| responsibly. Because of this acceleration, I don't think the age
| of AI will require less intelligent people. On the contrary,
| everything will likely become much more complex and abstract,
| because every knowledge worker (who wants to participate) will be
| empowered to do more, build more, and imagine more.
| pugio wrote:
| I've used AI for one of the best studying experiences I've had in
| a long time:
|
| 1. Dump the whole textbook into Gemini, along with various
| syllabi/learning goals.
|
| 2. (Carefully) Prompt it to create Anki flashcards to meet each
| goal.
|
| 3. Use Anki (duh).
|
| 4. Dump the day's flashcards into a ChatGPT session, turn on
| voice mode, and ask it to quiz me.
|
| Then I can go about my day answering questions. The best part is
| that if I don't understand something, or am having a hard time
| retaining some information, I can immediately ask it to explain -
| I can start a whole side tangent conversation deepening my
| understanding of the knowledge unit in the card, and then go
| right back to quizzing on the next card when I'm ready.
|
| It feels like a learning superpower.
| iteratethis wrote:
| I think there's ways for teachers to embrace AI in teaching.
|
| Let AI generate a short novel. The student is tasked to read it
| and criticize what's wrong with it. This requires focus and
| advanced reading comprehension.
|
| Show 4 AI-generated code solutions. Let the student explain which
| one is best and why.
|
| Show 10 AI-generated images and let art students analyze flaws.
|
| And so on.
| kaonwarb wrote:
| While recognizing the material downsides of education in the time
| of AI, I envy serious students who now have access to these
| systems. As an engineering undergrad at a research-focused
| institution a couple decades ago, I had a few classes taught by
| professors who appeared entirely uninterested in whether their
| students were comprehending the material or not. I would have
| given a lot for the ability to ask a modern frontier LLM to
| explain a concept to me in a different way when the original
| breezed-through, "obvious" approach didn't connect with me.
| walleeee wrote:
| > Students primarily use AI systems for creating (using
| information to learn something new)
|
| this is a smooth way to not say "cheat" in the first paragraph
| and to reframe creativity in a way that reflects positively on
| llm use. in fairness they then say
|
| > This raises questions about ensuring students don't offload
| critical cognitive tasks to AI systems.
|
| and later they report
|
| > nearly half (~47%) of student-AI conversations were Direct--
| that is, seeking answers or content with minimal engagement.
| Whereas many of these serve legitimate learning purposes (like
| asking conceptual questions or generating study guides), we did
| find concerning Direct conversation examples including: - Provide
| answers to machine learning multiple-choice questions - Provide
| direct answers to English language test questions - Rewrite
| marketing and business texts to avoid plagiarism detection
|
| kudos for addressing this head on. the problem here, and the
| reason these are not likely to be democratizing but rather wedge
| technologies, is not that they make grading harder or violate
| principles of higher education but that they can disable people
| who might otherwise learn something
___________________________________________________________________
(page generated 2025-04-09 23:00 UTC)